🔴 Alex Karp LIVE from AIPCon 10 | Alex Karp, Peter Zaffino, Chad Wahlquist, Sam Berry

You're watching TVPN.

Today is Thursday, June 4th, 2026. We are live from Palantir AIPCon. The temple of technology is back in LA.

The fortress of finance.

We will return to it, but it is also a state of mind. That's right. We are also sponsored by Ramp. Time is money. Save both. Easy to use corporate cards, bill pay, accounting, and a whole lot more all in one place. Big news from Ramp today. Massive fundraise. We're going to cover it in a little bit.

But first, you hear that?

We've got to talk Oh, is it still going?

I like it.

The ramp song's back. This was early days. We really talked about ramps so much, turned it into a song. Anyway, the topic of conversation in DC has— it's still in AI world, but instead of talking about approving models before they're released today, it's about the bio threat. Brandon Gorrell wrote in the TBP newsletter today, the great houses of AI have united behind the bio threat. There's actually a lot more to that because it was a big long list of signatories from AI, but also from the bio world and biotech and even startups. We've seen former guests of the show sign on. I'm excited to bring some of those folks back on the show in the coming weeks and hear more about this because I have this belief that as AI advanced, we got cyber because it was such a tight feedback loop, such a tight verifiable reward. Reinforcement learning works really well in that context. Bio has some similar characteristics.

And it was a very tangible Y2K style moment. Exactly. Where there was, it was a, let's just say, powerful business strategy.

Yeah, is it, yeah, it was like, is it over? You start thinking about the consequences of this and you don't need to get to AGI, super intelligence god. You can just have a really powerful tool that creates a new problem and that creates full employment for Nikesh Arora over at Palo Alto Networks, who we had a chance to talk to yesterday. And he's been very fortunate in implementing the solutions to the cybersecurity threats posed by new AI systems, some of the new AI capabilities that are rolling out. But bio might be next. And so it's exciting to see that the great houses of AI are uniting behind the bio threat. So let's take you through this. First, I'm gonna tell you about console.com. Console builds AI agents that automate 70% of IT, HR, and finance support, giving employees instant resolution for access requests and password resets. So in 1981, a group of researchers published the primary structure of the poliovirus genome in the journal Nature. So they were basically open sourcing the sequence for making polio, which just a few years earlier, polio I think was on the decline by 1981, but a very, very problematic virus. It's an RNA virus, meaning that its nucleobases or building blocks are A, C, G, U, if you're familiar with RNA, adenosine, cytosine, guanine, and uracil. Put more plainly, thanks Brandon Gorell, He says, "When the researchers published the primary structure of the poliovirus, they gave the world the literal sequence of poliovirus building blocks in order from start to finish. By the mid-20th century, before mass vaccination, polio was paralyzing and killing more than half a million people per year worldwide." So you have this pretty deadly virus killing more than half a million people per year worldwide, and you have just open-sourced it. What happens? So in 2002, researchers synthesized infectious poliovirus from its publicly available sequence data. So they didn't actually need any of the poliovirus RNA to start. They didn't need it on hand. They didn't need to, it's not like they took a little sample and they just cloned it up and made it bigger. They just took the data and they made the actual virus. So this is the shape of the threat. If there's a new virus or an existing virus or a forgotten about virus and you have the code to it, you can potentially print that RNA and then have the virus in your hands even if you don't have a sample. You weren't able to collect a sample. So instead, these researchers in 2002, they were able to take the published sequence, chemically synthesize short DNA fragments, assemble them into a full-length DNA copy of the poliovirus genome, and then use the DNA to make the viral RNA to fully recover the infectious virus. So in 2005, researchers used these same technologies to reconstruct the Spanish flu, a virus in 1918 that killed 675,000 Americans and had a 2% to 3% mortality rate among those infected. Very, very dangerous stuff. So basically these two reconstructed viruses showed that having a physical virus on hand was no longer necessary as source material to create viruses. All you needed was the blueprints. As long as you have the code, literally just like text in a text file, a bunch of ATGU, you can go and make this as long as you have the equipment on hand. But that is getting democratized as well. And that's what this AI letter is all about. So that's the situation that we're still in today, except now that we have AI, there are easier ways to potentially reconstruct DNA sequences that could create new viruses. So yesterday, Demis Hassabis, Sam Altman, Dario Ahmadi, Alex Wang, and dozens of other high-profile leaders across AI, tech policy, nucleic acid synthesis, and biotech signed an open letter called In Support of Mandatory Nucleic Acid Synthesis Screening and Record-Keeping. You might have seen it on the timeline. And at first glance, Brandon here assumed— and I assumed the same thing— assumed it was another press release from a frontier lab claiming it had just discovered new capabilities in one of its internal models that would ultimately lead to catastrophe. A lot of this, like, doom, fear-based marketing has been happening. So that was sort of the natural reaction. And that's what Brandon—

I think some people's reaction would be, were we not doing record keeping here already?

That's a great question. And Brandon actually did answer that. But it's not just a PR stunt, and it's not a new capability. They're not saying that the models can just create a novel virus, you know, one shot, like, that, that, that is Solved yet, uh, it's not there, but they see it as something that's coming down the pipe. And this letter is not this dangerous new capability, it's more asking the U.S. government to force nucleic acid synthesis companies to screen orders for sequences of concern. So, hey, somebody just ordered this, looks a lot like a virus, like, what are we doing here? Uh, you said that you were trying to treat cancer, or you said that you were, you know, trying to Make a new peptide and all of a sudden you're asking for poliovirus or something that looks like poliovirus. Like, let's dig into this. That's where they're going with that. And so they also need to verify the legitimacy of the customer and to keep a record of what they're sending and to whom. That's a crazy one that I'm sure you're like, wait, they weren't keeping records? They were a little bit. He gets into this. So he says, the reason the letter is coming out now is that the threat of nucleic acid synthesis sequencing getting into the wrong hands has been enhanced by AI. So anyone with an AI tool in the future could, in theory, if the models don't have safeguards on them, could create a sequence that then they go to a nucleic acid sequence company, get printed, send it to them, mix it up, boom, they got a virus. Not good. So most of the global nucleic acid synthesis industry has already signed up to do some of this. They started this in 2009. With what's called the International Gene Synthesis Consortium, and roughly 80% of commercial synthesis capacity worldwide is on board, but membership—

Okay, so 20% are still just hanging out. No, we're good.

80% of nuclear weapons are safely stored. Don't ask about the other 20%. That's kind of what this letter is getting at, because 80%, it was a good first effort, 2009, It's been 16, 17 years.

We haven't had—

There's a new reason to— Yeah, but there's a new reason to go further. Let's get that last 20%. That's what they're asking for. So membership is not a strong guarantee that they're actually screening or keeping records of their customers because it's voluntary. The 80% number is also self-reported, for example, and a bunch of other factors contribute to the relative flimsiness of the agreement. So it's not government-driven.

So you can opt into this program by just saying that you're opting into it. But then even the reporting, once you're opted in, is voluntary.

So I think the way this works is the International Gene Synthesis Consortium is probably a nonprofit NGO, non-governmental organization. And all the companies, they volunteer. 80% of commercial synthesis volume has opted into this. And then this organization, the International Gene Synthesis Consortium, they say, hey, we've looked at the market, and we're covering— about 80% has opted into this. We're on board with 80%., and the government isn't coming in and checking the records. They're not actually saying, okay, well, we have a different number because we're the government, and you have your this number, let's verify this number. It's self-reported by that organization, but there's no reason not to trust that organization necessarily. So, what else? Bunch of other factors contribute to the relative flimsiness of this agreement. HHS also has guidance in place around the issue, but again, it's voluntary, meaning that the possibility of bad actors getting their hands on dangerous nucleic acid sequences, at least from American companies, still cannot be ruled out. Overall, It's good to see industry leaders signing this letter and doubly refreshing that the letter is not yet another warning of apocalyptic AI doom, which I think the public has unfortunately come to expect from announcements like this. Hopefully the relevant legislators are paying attention and can make this happen in short order. So I thought that was a good, a good breakdown. And I agree with a lot of that. Andrew Curran also has some deep dive on this with some more of the signatories. He shares screenshots of all of these and it really is everyone. Yeah, Y Combinator, DeepMind, Microsoft, Interconnects AI, Harvard, tons of stuff. And then over in the nucleic acid synthesis industry, you have Twist Bioscience, ANSA, Emerald Cloud Lab, and Kathleen McMahon from Valthos is on here, former guest of the show. So good news, but obviously, and just an early step, this is just an open letter to the government. Saying, hey, we think you should— we want to support this. We think that the government should start thinking about this. Uh, the other news in the bio world—

yeah, I mean, the news is just that there's incredible momentum in biotech.

It feels like biotech, uh, after momentum, but not like volume, not scale yet, because you're looking at $3 trillion IPOs going out this year potentially. So much news in AI. Microns at a trillion. Every chip stock is, you know, in the hundreds of billions, trillions.

This is much smaller, but it's notable because biotech had been left for dead in some ways. We had a biotech investor on probably 14 months ago at this point who said, I don't even know. I mean, just looking at the returns so far, I don't know why you would invest in this asset class. But of course, every asset class kind of goes through that kind of phase, and clearly there's a lot of momentum.

And they should be. You would expect that biotech would be similarly power law driven, maybe not as extreme. But if you pull out SpaceX, OpenAI, Anthropic from the capital— yes, but I feel like the biotech community has a little bit more of like a culture of like base hits, doubles, triples, where they flip companies pretty, pretty frequently.

And yeah, we had that— didn't we have a guy on that had sold like 3 companies? We didn't have— yeah, we're doing billion-dollar exits.

Yep. And then he joined another company and sold it for $3 billion like the next day. Uh, and so that—

so anyways, you have Isomorphic Labs spun out of, uh, DeepMind, uh, Coinbase, uh, or not Coinbase, but Brian spun out or like founded NewLimit. Uh, you have RetroBots.

They just raised a new round. We're gonna get Jacob on the show as well for that. Um, Altos Labs, uh, from Jeff the Chad from Amazon, as this post puts it.

And, uh, Anthropic obviously acquired Coefficient Bio as well.

But Jensen and Larry Ellison at Oracle are also doing stuff. So there's a lot of activity. It's very fun. And I hope we're going to be able to cover this a lot more in the near future.

At what point do the— at what point does a Pfizer or Johnson Johnson start joining the press release economy of just coming— I'm not saying it'd be a good thing, but coming out and saying, we believe we're right at the—

Because there are partnerships all the time that happen. And they're always just like tucked a little bit deeper in the Wall Street Journal because AI is dominating and even private credit takes the front seat to the bio news. But there, there's a whole bunch of deal making going on anyway. There's other deal making going on in fintech. We're gonna talk about RAMP's raise today, but first I'm gonna tell you about Railway. Railway is the all-in-one intelligent cloud provider. Use your favorite agent to deploy web apps, servers, databases, and more while Railway automatically takes care of scaling, monitoring, and security.

So RAMP, what's going on in RAMP land?

$44 billion valuation. Oh, really, really solid traction. Just, you know, every 12, 18 months, sometimes much quicker. Sometimes they do 2 rounds in 2 weeks, but really solid progress. They raised $750 million at a $44 billion valuation. Last time we grew this fast, we were 1/20th of the size.

So this is actually accelerating. This is the most notable thing to me. Yeah. Lots of chatter on the timeline around other fintech valuations, you compare—

SaaSpocalypse.

Yeah.

Well, yeah, you compare them to SaaSpocalypse.

Ramp is now worth more than PayPal. PayPal has $32 billion of revenue, but PayPal certainly has, I would say, probably negative momentum, whereas Ramp has incredible momentum. And this is the standout line. They were 1/20th the size the last time they were growing this fast. Um, yeah, just really, really, really, really impressive execution. Yeah. Uh, and incredible opportunity still.

Yeah. Uh, so Eric, uh, took to the timeline, posted an essay about the third pillar comparing, uh, the previous eras of value creation, the two pillars, uh, people and vendors, dating back to 600 BCE. If you're not thinking in millennia, What are you doing here? Tokens emerge as the third pillar in 2026 AD, and he calls it the quadrillion token blind spot. Boil down 500 years of finance, and it's really just 3 questions: who spent what, was it worth it, what's the bill next month? I mean, people get caught up in all these crazy things. I mean, you see this with like marketing, I'm sure, and, uh, and ad buying, where people will do all these crazy analyses and ROI, ROAS, and all this other stuff, and like And it is always useful to zoom out and just be like, okay, we spent a bunch of money. Did the bank balance go up in this company? Yeah.

Yeah.

All personal and business finance at the end eventually comes down to, are we making more money than we're spending?

Yeah.

And I think Eric is right to dive super deep into token optimization and thinking about the tools that they're building, but then at the same time, don't get lost in the sauce. Actually zoom out and try and understand, like, what is the core value that you're delivering to your customer? Uh, it is answering that question. Um, so fantastic news over there. Let me tell you about the New York Stock Exchange. Want to change the world? Raise capital at the New York Stock Exchange. You gotta do it. It's my number one advice for founders these days. Um, uh, there's some other fundraising news. Sabe, the beanie BCI company, is getting preempted at $35 million at $500 million post. Uh, this is a leak from R4Rock. We'll see where it goes.

This is huge for you. Why? Because you are a beanie guy.

I do like beanies.

You love to throw on a beanie, a beanie in the morning.

Keeps it together. Yeah, yeah, I like a beanie. Uh, very, very—

I— it's interesting. I think that, uh, this format, of course, I'm sure they can adapt it to other types of hats. Yeah, but this format certainly maybe makes it harder to build momentum in places like California, at least Southern California, Arizona.

Be big amongst creative directors though.

Yeah, huge, huge, huge potential. Silver Lake. Silver Lake. Yeah. Every—

it's not too hard to change your beanie into a hat, a cowboy hat.

Yeah.

Like that's just extra leather around it. You could wrap the beanie in the, in the cowboy hat.

You can wear— here's what's interesting though. So R for Rock.

Yeah.

Usually it's pretty dialed. Pretty dialed.

Pretty dialed.

Uh, pretty dialed. It's almost like he has inside information. It's almost like he somehow got—

Yeah, but I mean, we've talked about the game theory of like, do— does he work at a real like tier 1 venture capital firm that's seeing— is it like, what's the benefit of leaking everything? Is he a lawyer that's seeing all the docs?

I mean, zero benefit for a lawyer, right?

Um, the rush of getting likes on the timeline is pretty universal. You're a lawyer, you're just like, I need a banger.

At a fund.

Okay.

For sure. Yeah. And I don't, I don't know anything else, but he's always taken the view that it can be helpful to the founder to build because a bunch of people are going to see this. Sure. That, that this didn't sort of land in their deal flow or land on their desk and they're going to reach out. Right. So it does create momentum. But, uh, can certainly be annoying for teams as well. Uh, this was notable though. So $200 million of, of LOI from B2B customers. And so very curious what the enterprise play is here. I don't know, but, uh, we can work on getting Rahul.

Does that mean like, uh, like through hospital networks or through like the healthcare system? Or is it like Mark Zuckerberg wants to go further? He wants to track the brainwaves of the employees?

Yeah, we're gonna, we're gonna track your screen.

We're also gonna check your brain. I mean, it could go either way because, like, you, you imagine, like, Neuralink has had a bunch of traction and a bunch of amazing— I saw, uh, Nolan, the, uh, the first patient, P0, uh, on Rogan talking about playing COD. Yeah, with the Neuralink. Amazing. Uh, and you can imagine that at a certain point, like, some sort of partnership—

they have multiple hat form factors. There we go, where hats come in. We're good. I was getting really hung up on the beanie. Yeah. And like, there's so many different enterprise or B2B contexts. You're in a warehouse in Dallas, Texas in the summer.

Yeah. You don't know. Maybe this $200 million LOI is from REI or Patagonia. You don't know who makes beanies. What's the—

Carhartt.

Carhartt makes a great beanie.

There we go.

You don't know any of this stuff. You're completely out to lunch on the beanie economy.

Beanie economy.

Anyway.

Beanie market map. Work on it. Let me tell you about public.com.

Public.com, investing for those who take it seriously. Stocks, options, bonds, crypto, treasuries, and more, all with great customer service.

They just launched a feature today that allows you to connect, uh, your favorite chat— yes— app to Public.

Yes. And more important than ever, because with Public you're going to be able to go and create the S&P 499 if you don't like SpaceX, or the S&P 1 if you love SpaceX. You can express your opinion about SpaceX however you want. Can you please help me build an Index for one company. Yes, index for one company or index for everything but one company. Uh, SpaceX is very, uh, divisive. People are extremely optimistic in certain camps, extremely pessimistic.

Goldman, Goldman, very optimistic.

What'd they say?

Uh, Goldman expects SpaceX's AI revenue to surge 100 times by 2030.

Huge.

Uh, big, big number. I looked at this title and I was thinking like, okay, what's Groq's actual revenue today if you take out—

Yeah, what is their AI revenue today? Is it just Groq subscriptions plus gross Groq tokens? Do you include X subscriptions? Do you include, uh, cloud vendor and Neo Cloud contracts? There's a bunch of different ways to measure it. The smaller the number, the easier it is to 100x. But we have seen other AI companies 100x revenues over 2 years, over 3 years, 4 years. Like, the 100x has become it's not a one-of-one scenario. Yeah, it's happened multiple times. And so, uh, we have seen these charts many times, and, uh, if, if they execute well, this is entirely possible. It is, it is extremely—

other, other notable data points from the roadshow. Yeah, they, the forecast anticipates SpaceX making about $360 billion of capital expenditures through 2028. Uh, Jensen somewhere fist pumping. Very excited about that number. Be a new hyperscaler. And anyways, very— should be unsurprising, but very aggressive. Yeah.

And yeah, the enterprise story is also live too. I saw the new NVIDIA Foundation model is also live. We'll have to go check it out and look at the model card soon, see how it's benchmarking. But we got to move on to benchmark because there's new news in the benchmark world. First, I'm going to tell you about MongoDB. What's the only thing faster than the AI market? Your business on MongoDB. Don't just build AI, own the data platform that powers it.

So, moment of silence.

Moment of silence.

Why is that? For, uh, the, the end of an era.

I guess, I guess they have been very focused for decades.

The last tier one that was a pure venture capital.

What did they get called again? Internet Boys or something? Soft Boys. There's some book about them that was very funny. But eBoys— eBoys is a hit piece of a book title.

No, it's a fantastic book.

But the subtitle makes up for it, and it's a fantastic book. And it's a very interesting story where they actually let a journalist come in and see how they—

eBoys: The True Story of the Six Tall Men Behind eBay, Webvan, and Other Online Markets.

Yeah, he clearly wrote the subtitle and was like, I gotta take the edge off of this. It's too glazy. I gotta, I gotta take it down a notch. And so he, uh, uh, he, he threw the e-boys in there.

But anyways, big moves from Benchmark. Clark has a scoop in the Journal. Benchmark has raised $2 billion across two new funds. Wow. And, uh, most notably their first ever dedicated growth fund.

Hmm. Do they hire anyone who has experienced growth investing? Who could possibly do growth investing there? Someone who's maybe like at Bond Capital and then Founders Fund, then maybe Kleiner, like someone with that pedigree.

Yeah, somebody with that kind of background, I think would be fantastic.

Pretty good for growth investing.

Now that you say that though, yeah, Ev Randall.

Ev Randall, that's right.

They did pick up Ev Randall.

They did pick him up. They were almost thinking two steps ahead there.

Are they building their fund strategy now, their entire platform strategy around Ev Randall?

Potentially, potentially. Anyway, uh, let me tell you about Shopify. Shopify is the commerce platform that grows with your business, lets you sell in seconds online, in store, on mobile, on social, on marketplaces, and now with AI agents. And we are very fortunate to be joined by Alex Karp in just a minute. He's coming in to speak with us at AIPCon here. We're going to bring him in just a minute.

While we wait, Austin-based podcaster Joe Rogan reportedly being considered for 60 Minutes.

60 Minutes. Uh, they're going to have to call it 200 Minutes. Bada-bam, because he records long podcasts in 60 minutes.

Hundreds.

It'll just be called Barry. If you're listening, put us in, put us in the ring. We're ready to go. You need tech correspondent, business correspondent, someone who can just chop it up for 60 minutes. We do 60 Minutes 3 times a day. We're ready to go. This is going to be light work for us, Barry. I'm ready. I'm ready. You could do 60 Minutes right now. You could do 60 Minutes tomorrow. You could do 60 Minutes You could do an extra 60 minutes easily when you're putting up— we're putting up 1,000 minutes a week. It's no problem.

We did consider that at one point early on. We thought, should we do basically a morning show? Oh yeah. Take a 2-hour break and come back and do another show.

Late night show. Yeah. Late night show maybe. Anyway, still, we have Alex Karp here with us. Here we go. Welcome to the show. Welcome back. Thank you so much for taking the time. We're going to have you grab these headset, these headphones, right? Not these. You can sit.

No, no, get close, get close.

Get in here.

We liked it last time the three of us were sitting here.

Oh, we can, we can put the giant—

let's put this up here, right here.

I'll sit, you can stand.

This always works. Makes me feel— yeah, yeah, get in here close. There we go. Good. Okay, uh, how is it going? How is AIPCon this time around? What's changed?

Um, well, we're in a phase. Yeah, all each one of these things like marks a time. First of all, You guys are even more baller, more successful.

Thank you.

Thank you. Some tendies in your pocket.

I think it might have been part 2. We got to say thank you.

You know, it's like you blew us up.

You're looking— biggest guest.

You're looking bigger and stronger.

Thank you.

Are you more attractive in your personal life now, randomly?

Well, here's what we're actually focused on. Dead hangs.

Yeah.

Yeah. So you came on last time, you said your dead hangs around like 5.

Oh no, it's— it's— well, it's plateaued in the last couple months at 5:30.

5:30.

Okay. So we— the thing is, like, people are going to hear that. They're going to think hanging on a bar, 5 minutes. You got to go and do it. The audience has to go try to do it. We've started doing it. We're still in the under— we're still between. Yeah, between 1 and 2 minutes, somewhere around a minute 30. You feel like your tendons are going to rip.

1:30 dead hang is respectable. Okay. 2 minutes is super elite.

It doesn't feel respectable when you have the 5-minute number.

You're looking at the time.

Strength matters. No, the thing is, I don't want to go into rabbit hole in training. The single biggest mistake people make is they try to hang every day. You need recovery. It's like anything else. So if you want to mimic and get progress, you just do what I do, which is once a week you hang as long as you can. Doesn't have to be super macho. And then that's your day. So like to say you can do 1:30, but multiple sets. No, no, you know, one day a week you do your maximum.

Max.

Wow. So like, let's say you could do 2 minutes.

Okay.

You try to do at least 1:30. You fight to get to 1:30, but you don't fight to get to 2 minutes.

Got it.

That's your dead hang day. And then you can basically fuck around the next day, do whatever you want. Don't overdo it, but you could do 2 times 1 minute with a long break. And then you just screw around, do less and less and less. 2 days before, if you're 2-minute dead hang, you do 4 times 15 seconds the day before you take off. And you do that, just keep doing that. The mistake people make is they hear my baller time.

They're like, "Fuck that guy." I mean, the mistake you're making is not doing a course. Course.

This could be a whole new—

it is a whole new revenue line of pounds.

You guys have it more for you guys. Yeah, call in the car. So, um, yeah, I mean, the dead hang is a— it's like— and also some of it's just genetic. Like, my other metrics are elite, but that— this is somehow alien territory. God-given gift.

What about, uh, what about breath hold underwater?

I don't do that, and I'm not sure I have it. Like, I grew up swimming, and I think I'm weaker at that. Like, I bet you I'd be in your guys' I, I'm a dive master. I can hold my breath for 3 minutes. I'm just saying, I'm just— yeah, so I think, yeah, you'd be like— I think you'd be crushing me on that, honestly. But you have like the lung capacity.

That's true, that's true.

I mean, like, yeah, you got like—

yes, I'm not moving, I'm not using any oxygen.

It's like you're like, like you're like a, like a, like a whale floating out there into the ocean waiting to surface. True. So, you know, I say— I'll tell you the difference, uh, and there's this— oh God, they're always minding me out there. But like, okay When we first met, it was like, AI, it may be real. Then I would say somehow until about 2 weeks ago, there was like a, holy fuck, this is real, but somehow it's not working, but we're not allowed to say it publicly because we'll look stupid. And then there's a lot of investor hype. There still is investors printing 10Ds. So you have the investors on one side. I think people realize it's real, but it's like you have the whole— token maxing, and people are on to that. And then so there's a whole value lecture there. You have a political situation where people who do not understand basic economics are winning the political argument. So we could talk about where—

Yeah, there's a lot here. Let's break it up. Let's start with the token maxing thing. Let's start with what's real. How are you actually thinking about deploying AI.

First of all, what is Palantir's philosophy around token consumption?

Well, okay, we have a product that will allow you to— I mean, internally it's called something, but externally, but really we call it the demasturbatory, like get off masturbation thing internally. Sure. It's like people are just sitting there all day, kind of like a porn addiction. And enterprises are like, okay, we knew this We believe this will create value, but we cannot have people just like some people—

Checking the weather with it, just rearranging deck chairs on their personal Titanic.

It is literally like porn. People are like full on—

It feels so—

Yeah, tool-shaped objects.

Tool-shaped objects you're looking at more than you want. You hope no one notices. You're kind of before dinner.

It feels productive to have every email classified with tags.

Here's what it comes down to. I mean, sometimes they can be solved purely with money and just spending more, but very often—

Actually, I think it's the opposite. So just to give you a weird analogy.

No, and I was going to say very often it's more the opposite where it's about figuring out the right way to do something and then you can use capital to fuel that process.

Let me give you a thing that's too generous for you guys. Okay. It's taste plus money. Okay. And there is no, like AI, like if you look at, like pick any issue you wanna talk about. Token maxing. What's going on with deploy codes? Are other people gonna build ontologies? Why does our political class not understand AI, especially in Europe? It's like, yes, because all these things can be scaled in a very valuable but largely going to commodified way. But you can't scale the taste of like, what is the business problem you want to have to solve and need to solve? At the end of the day, whether it's the Ukrainians fighting, the Israelis, commercial entities, there's somebody sitting there who's like, okay, but this problem is valuable. This problem isn't. And once that value, that problem always has, that problem almost always, but not always has attributes. So there are some problems you could solve with this. I want to write a report on GDP growth in China, right? OK. But if it's a problem that requires a knowledge store, like I want to understand the specialized way I underwrite— we're going to have a guest here— I want to understand the specialized way I drill for oil and gas that's both legal, ethical, and reduces the cost of production. I want to change the supply chain of my industry, whether that's military or whether that's building boxes or whether that's cars. These things require actual precise ongoing processes. They are enhanced by large language models. They are not replaced by large language models. And then you get to security issues. Like for us, like the whole Mythos things is just a boon because like, yeah, we could take any model, their model, OpenAI model, open model. We can identify, we can now identify vulnerabilities at like 10, 100x. Yeah, but then who patches them? How do you patch them on-prem? How do you patch them on-prem so that your specialized knowledge stays on-prem? Like if you're any business or intel service, a lot of these things are very similar. Like you're not putting your classified data in a public cloud. Same thing if you're like, you have a special way of farming soybeans. Well, you're not. So it's like, how do you have, how do you get, so all these problems are exposed identified. And then you always have a thing of where's the charisma, which people really underestimate. And it's not global. There's no global charisma now. So right now, the large language models are very— frontier companies are super charismatic with investors. I'll give you some news. They're super not charismatic with enterprises and the people. In a way—

No way, even with enterprise?

No, no. I mean, it's—

Because I understand what the people are like.

No, no, the enterprise people, I have a secret. I have a secret. Like, every company has a secret way of selling. You know what my secret way of selling is? Don't even call. Don't come talk to us. There's a frontier company. Go spend 2 days with them. And if you're lucky, after you're done, I'll let you in my door.

They're like clamoring.

They're like, hey, I'll take your bad brand, which we have a great brand enterprise. But it's like secret knowledge because the investors love this. They're like, hey, my stocks are all up. Everything's up. I mean, Palantir has done very well. But it's like— and you guys are doing very well, I imagine, right? And it's OK. But I'll tell you what. You go down the street. You talk to a Marine. You talk to a bus driver. You talk to the person who owns the bus driving company. They are not happy. They do not like these people. They're tired of people token maxing. It looks like masturbation that's cost them money. And honestly, then you have something we're not allowed to talk about in this country, likability. Like, at Palantir, I think we have like 50, 100 million global fans. We have like 5 million people that wake up in the morning literally calling me Satan. I didn't know I had that kind of warm hand. But, you know, it's like, that's what they believe. And they really believe it.

Okay.

What people are not allowed to really address is like, we have fans and enemies.

Yeah. Yeah. You're polarizing.

Yeah. We're polarizing, which means both sides. These people have one side. They're just— it is. So it's like, and it's like, it's a really—

Social media companies too have the same problem.

Yeah.

Everyone uses them, but no one likes them. Yeah.

But then they also live in a circle and that circle's printing money. Yep. So it's like, you know, when you look in the mirror and you just printed a lot of money, you look pretty fresh.

Is part of it, is part of it that some element of the technology, let's just say LLMs, is so magical that the companies involved, that the companies that are making and selling frontier intelligence can be bad at a bunch of other things and still grow?

Well, no, no, no. They are magical at a certain kind of thing, allowing you to write, for example, code. Now that code doesn't, can't be used as a knowledge store. So if you look at code in like 3 different ways, like just using Palantir as a model, we have code that's basically infrastructure. So what are the Ukrainians using? What is the Department of War using? What do a lot of our enterprises, we call that primitives. It's basically hard-coded things that, that understand the world the way you do. It would take millions of technical hours and an understanding of all these enterprises to do it. So it's much more like, how do you build a steel beam? Then you have code that is written by FDEs. Okay, so that's kind of managed. The reason why FDEs work, the secret is it's actually managed in something that we as a product. So you're writing to a code base, we're managing that, we're increasing our product. It's not just random people writing. Then you have, let's call it free code. That free code is, that's magical. Like you can do it very quickly. It's almost right. It doesn't have to be exact. Dashboards. Dashboards, financial stuff.

Little flows.

Probabilistic stuff where you'd have to get—

One-off analysis.

It was magical. By the way, it's magical. It not only creates, and it's magical in a way. I know people don't like the porn thing, but it's also addicting. It's like, you know it's not good for you, but it may lead to damage.

One more dashboard.

One more time. It can't hurt that much. I know my doctor says I shouldn't do it. But it's like that, right? And you just keep going. And if you're involved in that thing, you're also making money. And then last not least, in certain circles, like if you want to be a researcher or you believe— essentially, it's a religion. And one of the things that's very charismatic, especially to people who've never had a religion, because all of a sudden that hole in your heart that was yearning for, I don't know, I would say, you know, a established religion, Judaism, Christianity, Islam, is like being filled. And all the answers are there. But it is very, very successful at doing things that a company has to do. But it is not actually solving the problem that enterprises are. Now, it can solve them. That's the trick. It's not binary. It's not like you can't say they're not valuing. They're totally putting our business on steroids. Like without LLMs, Nobody would be talking about our ontology, about Apollo managing secure exploits, about our ability to manage enterprise essentially, turning all these companies into FTEs. These deploy codes, we love them because now every company wants a deploy code. You know how you do that? You re-platform on Palantir. And it actually works. It's not somebody with no taste who's never done enterprise, who has no earthly clue how these things work, who's done something else and is just imagining they know how to do it. Right.

Yeah. As part, as part of this moment, quite entertaining for you because you guys have been working on understanding businesses at a deep fundamental level, creating, you guys have effectively been doing the work that it, that people are promising AI could do for 20 years now, but actually doing it, finding all the really rough, finding all the really rough edges. And, and not— and being at a point where you don't have to oversell the technology, you can sell both things. But now there's maybe— here we go. We got it together. Now there's maybe—

oh, it's the wrong side. That's why.

Flip it around.

All right. Go.

Dyslexic.

Yeah. There you go. There you go.

Living the brand.

Hopefully we got that. I don't know. All this stuff was recording.

I trailed off, but— no, I know. Is part of it entertaining to you that It feels like Palantir has always been in some ways not had competitors because there's nobody with Alex Karp running a company that does what Palantir does besides Palantir. But at the same time, there's been tens of billions of dollars deployed now to effectively do what Palantir does, but just selling the intelligence part, not selling all the underlying kind of infrastructure that—

well, they're doing two things. They're selling— they're trying to sell the intelligence part, and they're trying to pretend if you just hire a bunch of people and let them run around their FTs. Now, the, the very cool thing is when you've been in your basement doing your thing and everyone kind of use it as the freak show, it, it's really interesting and, and great to have adoption. The pretty ironic thing is half the people adopting now don't even know they're copying. But now the copying thing, it helps and hurts. Where it hurts is in the beginning it puts clutter in the market.

Yeah.

Um, and there's, there's no doubt about it. Where it helps— and then we saw this with defense tech, honestly. So like, in defense tech, we were the only people, we were the first people, despite what— I, I love these, honestly, other podcasters, they're interviewing people who are parroting things I said 20 years ago, they don't know it, and it's like, oh, that's so insightful.

It's like, yeah, of course it's insightful.

Karp I said it 25 years ago, uh, and like, but it says— but so that kind of— that part is super weird. But, uh, and like, but, um, but it's— but what really happens when we see is like it expands the market. So like in defense tech, we would not be doing this well in just purely in government unless there weren't 50 companies that were doing similar things, because then the people are like, okay, first of all, you view it as like off-balance sheet sales resources where other people are basically— Well, it's off— well, no, that, that's the large— like, it's— they do two things. They increase the size of the market because de facto nobody wants to find it— underwrite a market where there's only one person.

Sure.

So like, if you're the one person, the percentage of the defense budget you can get is much smaller. And two, they set up a comparator. It's like, you know, you may not like the freak show, okay, if you like, but have you noticed the people who are serious buy it? And then it changed. And then 3, it changes the standard. Now, what you're seeing now is like that times 100x. And it does change recruiting, retention, and how you build a company. And we're always thinking, you have to think about how to— being dyslexic, huge advantage there, because you don't have a playbook. And now that you need things to shift, and we're doing that. The central thing though, that it just cannot be developed even if you understood the playbook. A lot of these things appear like, it's like LLM code appears like Palantir code, but isn't for a deployed thing. Appears like Palantir, it isn't ontology. You could theoretically copy parts of it, but they're essentially structures that are built deep into organizations that we own. And by the way, take you 3 years, and then 3 years we're in a completely different world. But there is this magical thing called taste. Like, in the end of the day, the reason why you guys have done so well, it's— of course there's aptitude and diligence and showing up and all those things. Yeah, but you have to be able to differentiate between two people who are in business, one of whom is saying something that sounds weird that is insightful, one of whom is parroting something that sounds weird and that's all they're doing. And a lot of people— very few people can do that. And you have the same thing, like, the enterprises that succeed, there is a taste arbiter. And at Palantir, we have taste in every product, taste in every deployment, taste in every casting. Who puts the people there? How do you put them there? How do you organize the thing? Our ontology then does that technically. How do you manage the whole org with taste? Who should be in charge? What data sets should come in? What are the ways in which you protect? What should you push into the public crowd? What should be on-prem? What should, I mean, leaving aside the law and like wars, war ethics, What do you want to protect? What should you protect? What should you not protect? Because quite frankly, you want that to be out there so you can get more data. All those things are arbitrated by taste, and then you have to have the credibility of having taste. That's a real problem for a lot of these places, because they don't have— they're popular with their friends. They really don't understand how unpopular they are in enterprise. They think it's like, oh yeah, it's like the way I think I have a problem with professors at Columbia. It's like, no, it's a real problem. They think I'm Satan, and it's like, I think You know, we grew up in the same community. Let's talk about Heidegger. They're like, they don't want to talk about either. So it's like, it's like, yeah. And so that's just a, it's a weird thing. It's going to be a super— the one thing I would say for anyone listening, if you're listening to this and you're chillaxing and not active, I'm not saying you have to agree with me politically or anything. Yeah. They're the, like, partly because of this dynamic and very self-inflicted because I, I'll tell you, I can't name names. I called many of the titans of this world and started this 6 months ago. Every couple days, we're going to be nationalized.

You call them every couple days?

Some of them, I'm like, yeah, we're going to be— I mean, honestly, they're like the bad— they find me very entertaining. I'm not sure. So they call because, yeah, it's like, oh yeah, this is going to be entertaining. You're going to pick up. Yeah. So any case, I've been telling them for 6 months, we're going to be nationalized. We're going to be nationalized. And they're like, why would anyone nationalize? Never happened in America. It's never— why would anyone nationalize us? We're so likable. We're creating so much value. Like, okay, I'm not going to debate that. I know how likable I am. I'm not going to tell you how likable you are, but I am telling you, and you know, the momentum on this is on the side of people in national— and we don't get our act together and figure out ways we can say, hey, look, there are problems here we're going to deal with These things are not going to— yes, they are going to create opportunities. You have to talk openly about how these things are valuable because we have adversaries. You can't just say these— all that stuff. So the primary risk, honestly, to Palantir and a lot of these other countries is— and then it's going to be nationalized before nationalized, it's going to be regulated by people who don't understand this. And now they'll tell you in private, I'm working on this, I'm da da da, and this lobbyist is like, not going to work. So like, that's something like, if you're listening to this and you're like, Look, you know, you don't have to agree with me on all my proclamations. I got a lot of— by the way, there's some people who think I'm saying we should have a draft. Too lazy to read. I'm just saying we should— like, in a world where everything is changing, everything is changing, don't we have to find some communal structure to remember we're American? You don't like my— you don't like my idea of like we all do a week in the park? Great, come up with some other idea. We can have no idea, you know? And like, and then they're like, well, I'm saying I do not want a draft, just to be explicit. They're like, Oh, that's pro-war. No, honestly, you know what? Most of our wars are fought because no working-class person is making a decision. You start making sure everyone is involved in everything, I'll see you have few wars we fight. It's actually the anti-war position. But any case, disagree with everything. If you're— we have on the right and on the left people, uh, people who have no earthly clue what they're talking about, right and left. All they're talking about is how much they hate us. And those of us who are sensible in the middle You know, too many of us are chillaxing. Like, oh, like, nationalization, it can't happen. America would never do that. Sleepwalking into— and you guys have tendies to protect now. You guys should be on the front line of this. Like, you got full— oh, I'm sorry, I have a, I have a full-on, very impressive, uh, corporate leader coming on. Yes.

So I got it. I got it. Last question, last question if we have time. Uh, how are your conversations going with Fortune 500 CEOs around headcount planning? There's been so many layoffs this last year that people were saying, hey, we're getting so much out of AI, we're able to, you know, cut back here or there. People inside tech often know, like, these— maybe there's just reduction because there needs to be a reduction. The org got bloated. Maybe they do need to fund some AI.

Declining business model.

Like, yeah, the business just doesn't have momentum. But how are those conversations going What does it look like?

Like, by the way, I talked to Fortune 500 companies, I talked to unions, I talked to soldiers, I talked to fire— if you upscale somebody, they're more valuable. Sure. And like all these, whether it's people working on batteries, people driving trucks, people, corporate leaders. And again, this is where I think we have to be very careful to be more disciplined on the corporate side. Like if you run around saying AI allowed you to fire two-thirds of your workforce and you did it, because maybe your competitor's kicking your ass. Yeah, that could— that is a really— like, you might as well just go sign up for Bernie Sanders Manifesto. And part of the thing is they really believe that can't happen. So they're free riding on the fact that it could. Like, we have— and it just cannot work anymore. These things are very, very explosive. The American people sense that there is something dangerous here. And when people are playing with that fire, it's like they assume the fire won't burn their hands. That's not the world we're in. That fire is going to consume us. And what we see, again, the warfighting example is just the most neutral, not for everybody, but like the soldiers at the bottom have gotten much more valuable. And I don't even just mean the special operators, which obviously they're in a different league, but like every— the people doing a lot of the operations now are doing our prac, they're high school vocationally trained. You see this everywhere. The modern enterprise is going to have like we have a true like He very, very, very smart, uh, person coming on. And it's like, you're gonna have a very smart executive. He's much better at hiding it than I would be if I were him, but that's— you can talk to him about that. But, um, uh, um, and, uh, and then very talented creative people with taste all up and down the stack. Any case, I think this is time for me to—

I think this is time. We, uh, thank you so much.

Yeah, great to catch up.

It was fun. Oh, First.

Oh, they want me to stay for 2 minutes or what? Oh, I'm only going to stay— look, but he's got to be the star.

Uh, the other headset.

Put it— put it in here. Yeah, and I'm just gonna— I'm gonna take off after a minute.

Here, here. And why don't you— here, put that headset on.

Carl, why don't you introduce our guest?

Microphone on the left.

Well, I— yeah, you know what, he's one of the smarter people in, in business, um, has developed unique ways to underwrite that did not involve firing people. And someone, and someone I admire.

Thanks, Alex.

With that, I'm gonna let you guys go. Make sure to tell them that the ontology powers it.

It's everything. Always selling.

Hey, fantastic.

Thanks for coming on the show. It's great to meet you.

Uh, yeah, please, uh, kick us off with like a bit of a more formal introduction.

Yes, I'm Peter Zaffino. I'm the executive chairman, as effective on Monday, of AIG. Congratulations. Chairman and CEO, and have, you know, worked with the company for 9 years to help transform it. It was in a place where underwriting profitability was challenging, operations were challenging, data was challenging, capital was challenging. So, you know, had a great team of people with me to transform the So give us the—

give us the shape of the business in terms of the different business lines, the different products, the international footprint, the workforce. Like, what— give us the scope and the scale here.

Global company, yeah, with a little bit of a unique footprint. We're 50% international, 50% North America, but our second largest country after US is Japan. Oh, um, we have a big business in India.

Okay.

Yeah.

Um, and then we have a very big business in, in the UK. Uh, we do complicated risks. So you could think about what's happening in the Middle East now with shipping, marine, energy. We're heavily involved in that.

So something where there's not an existing futures contract that a company can just go and hedge. It's not, no, I'm going to buy some oil futures because I, I fly planes around and I know I'm going to need diesel fuel in a couple months, and so I'm going to hedge that out. This is for more complex risks.

It's for more complex risks. And, you know, you think about the largest you know, sort of customers in the world and, you know, big oil companies, you know, Fortune 500 companies. But we also have a personal insurance business which will cover things like accident, health. Yeah. That are distribution to consumers.

Yeah.

So we have a real balance.

Part of that feels like if you're talking about insuring a Fortune 500 company against a geopolitical risk, that feels like a meeting that takes place in a boardroom. It feels like there's a lot of folks with a lot of trust built up over years to understand each other's businesses. Uh, but then there's probably a lot of other underwriting happening and teams, uh, putting together comps and spreadsheets and data. And I want to know about the intersection there. It feels like the business is— and I don't know if it ever will be— just one-click checkout for, for insurance products for Fortune 500 companies. Uh, but what, what is the interface between the quantitative, the qualitative, the relationship, and the data? And then how is that changing?

So the quantitative, you have to start at the portfolio level. Okay. And you want as much data as you possibly can to look at deterministic modeling, probabilistic, and then stochastic. And I think once you understand like your mean and you understand the standard deviation around that, then you have to apply it to, you know, sort of the widgets, which is each policy throughout, you know, the globe as well as ways in which you structure insurance.

So for us— You can't look at an individual policy in in isolation. You're, you're managing portfolio risk, risk to the entire firm, and, and that's something that's happening probably 24/7, I imagine.

It's hard, and that's what led me to Alex Karp. Yeah, you know, it's hard to get the aggregation done in anything that looks like real time. It's usually static. It can be 30, 60, 90 days, and your portfolio could change. I mean, it's not going to change dramatically, but having the ability to, you know, sort of assess risk and use the quantitative data to make better decisions on a daily basis is the aspiration of the way the company property's going.

Yeah, uh, take us back to your first meeting with Karp. Curious, uh, what the experience was like. It's a unique individual.

Call you?

Yeah, no, um, I was actually introduced by a board member many years ago, and it was really in this pursuit of, um, not necessarily Foundry or AIP or Ontology— that's where it led us— but it was more on sort of the quantitative ways in which I was looking at the portfolio Could he help me think through computing and could he help me think through sort of portfolio optimization? And I just got more and more intrigued. I mean, you see the brain. I mean, he just thinks about things. He doesn't hold back.

I mean, so he's—

so I always knew where he stood with me and with AIG, but just developed a very strong trusting relationship. And they're such a tremendous partner that we're able to iterate with them almost like no other company because we do things in 90-day increments because going out like a year or 2 years is, is too static. And so we actually build our relationship on 90-day goals. Okay. And that's been incredibly effective.

What is, uh, you know, a lot of the AI companies talking about scaling laws, exponential growth and token production or even revenue in many cases. But what's growing exponentially in your business? Are you bringing exponentially more data into the platform every year, exponentially more compute resources, teams, number of policies. Like, what is the— what is the thing that's experiencing a boom right now?

The most important part, I believe, in terms of business is that you have to have a business solution you're trying to solve. So for us, it was more data.

Yeah.

Um, better data. Yeah. And then reduce cycle time. So in other words, like, when we get the data that comes in from our distribution partners. How fast can we get it with higher quality data and more data to the underwriter to make decisions? Got it. Um, and then how do we actually make really—

what's, what's an example of distribution partner in this context?

So it'd be like an insurance broker or insurance agent, um, or, you know, someone who has their client is a customer product, effectively.

Exactly.

Okay.

Yes. Yeah, that makes sense. Um, what else? Jordi, do you have something?

Uh, Where was I going to go? The—

I don't want to cut you off. Alex wants to cover Anthology.

So yeah, we'll get there. So there's been— we primarily, I mean, we at least started covering early-stage startups. There's been a debate in our kind of little sub-industry right now around a bunch of new insurance-focused startups that are growing incredibly quickly. And there's a debate going on as one maybe AI makes it more possible to underwrite risk. And if you can do that well, grow very quickly. The other side says, hey, if you're hyperscaling an insurance company, maybe you don't want to work with a company that is going through that hyper—

The iron law of the universe.

Yes, yes. Maybe. What goes up fast must come down fast. But yeah, talk about Talk about what AI has actually enabled, where you're excited about it, where it's failing broadly, maybe where it's overhyped, and you can, I guess, tie that into everything you built with Palantir.

There's never been a time, in my opinion, whether it was, you know, introduction to fintech, insurtech, how to use algorithms, how to build data lakes and repositories for data, there's never been a time in, in my professional career— so it's 35 years in big companies— yeah, that I've seen the ability to change how an organization actually runs itself. And that can come from big companies like Palantir or Google, or it could come from, uh, you know, companies that are being funded by venture and have a very specific niche that can be, you know, additive to the organization. And what I think is happening, we talked about the sort of data ingestion portion, getting that into a digital workflow, using large language models to extract more data from what comes in, but also helping underwriters make decisions that are, you know, more comprehensive. You also have the ability in the way in which you service customers to be much better through the use of AI. I think companies generally, uh, my observations are struggling with the orchestration of how you actually drive agents, people, and data into an organization. And once that is solved, and it's certainly on its way, capabilities are there, um, then you start to think about the entire end-to-end chain being very different.

Yeah.

What I think about Palantir, while they've been such a critical partner, is one, as we evolve together, but in that data ingestion, to be able to take structured, unstructured text, all sorts of data, and get into a workflow in a fraction of the time helps us on the things I try to achieve. Like, we have now data that we probably wouldn't have used before because it wasn't good, or we couldn't translate it, couldn't get it into the digital workflow. Um, and then we start to build out an ontology, and I really do think it's incredibly important. If there's one thing I look at for our organization certainly the advancements of LLMs, their ability to do things more autonomously now where we started with the binary gen AI, now we're into agentic AI where it can just do things autonomously for so much longer without the ontology of actually building like what the sort of digital twin of your business looks like, where you take it and how you evolve it becomes very challenging. So we've been able to do things with Palantir. I'll use the ontology example again. We did the full, ontology of AIG, and then we went to look at an acquisition called Everest, which had about $2 billion of premium. We got Palantir in to work with our team. We could build an ontology of Everest's portfolio on top of ours in 4 days. And quite frankly, what we started to learn again about that evolution is that you always relied on data lakes or global data repositories. What we found is that we could get you know, sort of foundry and start to build out this ontology with going to the admin platforms. All of a sudden, these repositories and the central places of getting data and make sure it's scrubbed wasn't as relevant. So I think we continue to advance that in, in the way in which we are looking at our business.

I have a lot— I have one last question, um, just on the actual change management, the organization, like how the office feels. What— how did you go about actually working with Palantir? Do you set up your own internal Palantir workforce who sits alongside FDEs? Do you let Palantir come in and plug in like one person per team that you have set up? Like, was there a best practice? Did you go with the best practice? Like, what was the actual like experience of deploying the forward deployed engineers? They get deployed into the organization. That's got to be a unique situation?

First is making sure Alex and then, you know, two of the senior executives, Ryan and Ted, that everybody knows what we're trying to do together. So we start there. Then we wanted to embed the engineers with our team. So if we had a business leader that was trying to drive the underwriting output, you'd have, you know, technology from AIG, you would have some of the change management, but you have the engineers sitting there with our teams throughout the entire process because the iteration is really important in terms of translating what you're trying to achieve from the business side engineers actually helping us think through the application of some of the LLMs or ways in which we could circumvent some of the things that we were doing.

Yeah, that makes sense. Jordi, anything else?

No.

Well, yes. Insurance, I think, would be the most important topic. No, no.

If we do have a second, I was not sure on timing. How are you thinking about workforce planning? Asked Karp about this and he said to ask you. Or token budgets. That's interesting too. We've stayed, as you've had the this wave of AI layoffs, we've been over and over and over reminded people that if you have an individual and you give them more capability, you make them more productive, you make them more efficient. A thriving business will want to hire more people, right? Because you can get more out of every individual. And so we've tried to remind people of that over and over and over, as, you know, companies that oftentimes are, you know, underperforming or bloated for whatever reason., but what's your kind of philosophy around, uh, hiring, headcount planning, uh, RIFs, all that stuff in this kind of, uh, new era?

We've been focusing on— I heard Alex at the tail end and I agree with him. So we're focusing on growth, uh, we're focusing on reskilling and actually training, um, our employees to be in a different part of the workflow. Now you would do this, I believe, in all of this you have to still have great end-to-end process. And so things that have been— the humans been in LLM trained how to do things like outside of the normal workflow has to— you have to get rid of that. I mean, so I think that's just normal business. Yeah. Um, but you know, our aspiration is not to implement, you know, AI or anything that we're doing with our partners to eliminate jobs. I mean, it's about growth, reskilling, and finding ways in different markets to have exponential growth and opportunity and having a lot more insight in the business that we run.

That's a great optimistic vision. I love it. Thank you so much for taking the time to come chat with us.

Thanks for having me.

Have a great rest of your time. Thanks. And up next, next we have Chad Wallquist. First, I'm going to tell you about CrowdStrike. Your business is AI. Their business is securing it. CrowdStrike secures AI and stops breaches. Welcome to the show. How are you doing, Chad?

Great overcoat.

That's a new one.

It's a popular one.

That's an Eliano special.

It is. Oh yeah, he is the master.

Giving us a run for our money.

Yeah, it's fantastic. Uh, anyway, kick us off with an introduction on yourself, how you fit into Palantir, a little bit of backstory. I'm sure we have a ton of questions to run through.

Uh, first, how often do you guys do these things? Because it feels like this feels like an annual— it feels like an annual event. Yeah, yeah, but, but we're getting the call every 3 months. Karp talks about, you know, manipulating time, you know, uh, a quarter at Palantir is like a day.

Yeah, a year somewhere else. That is exactly—

so that it kind of makes sense.

Yeah, I'm like actually 23. The time warp is real. So we do these quarterly. So I'm a forward deployed architect technically. I do what is needed. And so doing the needful is kind of the Palantir way. It's like there's no job below me. And so no matter if I'm out on the edge with customers, I'm talking to executives, explaining the ontology, doing YouTube videos. That's all what I'm doing. So really the goal is how do we help people decomp problems differently and apply the technology? Anyways.

Can AI do decomp?

Yes.

Okay, unpack that because that feels like the secret sauce. That feels like the special thing about Palantir is actually being able to bring someone in who understands an organization. I think a lot of people see AI tools. A lot of people see AI tools. No, a lot of people see AI tools and they, and they think, uh, okay, very defined workflow, input, output. But now instead of just math that Python can deal with, you can deal with some text, and that's great. But decomp to me has always felt less like, let's go into your HR system and understand the basic job description. Like, oh, someone uploaded this resume versus, oh, Steve actually does this completely outside of that system. And marketing has two, two platforms for this thing, and engineering has three systems for CAD files. And they're all the kludges that have built up over decades, sometimes hundreds of years for some of these organizations. Like that's what was so special about the forward deployed engineer program, the Palantir model. I'm surprised to hear you say AI can do it at all. It feels like the final boss.

Well, this is where the really the Palantir thesis is humans and AI working together. And so the way we think about this is modeling our business process. We heard some other people talking about this of modeling my business process into the ontology because the LLMs just don't necessarily have a worldview or world model of your business and your operations. The ontology provides that. Okay. And so when we talk about decomp, this is really about actually now I make more data computable as well. So we think about LLMs on the agents and I'm interacting with, but also we use LLMs to make more data computable and then model that in the ontology of how things are really working. And so what we are actually doing a lot of times now is building out that worldview and then running multiple agents over this actually being combative towards each other. Right. And so actually working against each other and having critiques. And so after you do that, you can also then give the human, human-in-the-loop feedback about this and iterate on this. And also what we find is that's really a scaling mechanism. It's like a new power tool, right? I think you guys were just talking about this, the kind of the perspective around jobs and all this stuff. It's like when you gave carpenters power tools, there weren't less carpenters, there were more. I could do more with it. It's an empowering thing.

Yeah. So, uh, how often— like, I, I, I'm interested in the, uh, like, the pie-in-the-sky Palantir pitch— understand your entire business, run your entire business on Palantir— and then some of the nitty-gritty where sometimes, like, the low-hanging fruit is like, wait, there's a— like, there's someone's job to just, like, take a form and type it into a sheet? Like, we have— we've had image recognition for a long time. Let's actually go implement that and get that into a database, get that into the ontology, get that into Palantir so then we can start building on top of it. And it feels like there might be a tension there. Obviously both processes are speeding up, but how do you sort of like keep the project centered around the big goal while still chopping wood on all the things that actually need to happen?

Yeah, I think this comes back to the forward deployed piece and like what do we deliver outcomes and we work backwards from that rather than, hey, I have this data, I'm going to build a data warehouse and then I'll build reports because all my data is in one place. That's the field of of dreams and no one shows up. Right? And so really when we decomp things and work backwards from that, you know, the simple things like the form filling out, there's a lot of that. Now the one approach that we see a lot is enterprise software is going to force you into their box. Sure. Right? You go fit into this box.

Yeah.

Well then, you know, okay, did I take away the special sauce, which was my company? Because people were doing these, all these kind of amalgamations. Hey, 40 ways to do a PO. Well, maybe it is okay to do 40 ways, but my software can't handle it and it's fragmented, right? And so there's, there's actually a middle ground because, you know, for a long time customization was kind of a four-letter word, right? You know, no one wanted to do that. And I think that's where we think about malleable software. Actually, how do we help you be more different, not more similar?

Interesting.

And that's so that when we decomp problems, thinking about not only the, the kind of the quantitative piece, but the qualitative piece and the people and process around this, how do we enable those people to do the things that made them special?

Is software getting more malleable? Because I can look at it two ways. I can look at one, you know, obviously AI agents are incredible at coding. They can run, they can make changes very, very quickly that would take you a day in just a few minutes. At the same time, I see, you know, so many screenshots of people saying, I implemented this feature again. The GitHub is plus a million lines of code. And at a certain point, like the context window is growing as fast as the code generation is growing.— there's a— I'm a believer in the answer to bad slop is good slop and more slop maybe. But what are you actually seeing on the malleability of software? Because sometimes the most malleable software in the past has been, oh, well, there was a really incredible engineer who figured out this problem and baked it down to a 2,000-line repo. And you can actually just put it in your own context window. So it becomes more malleable and you can use it as a building block. And that feels like that's going away and I want to make sure that we've, that we're ready for when it goes away and it remains malleable.

Well, I think what's missing is the, the malleable enterprise scaffolding that you, and that's what we think about the ontology and Foundry and the platform and then Apollo that allows us to go deploy these changes.

Sure.

So it gives us the right amount of structure but the right amount of freedom. So I think that's the balance we try to find is that malleability in the middle where we can actually scale, we can enable people to do things differently while still creating enterprise-grade, robust, secure, scalable software. And so it's actually a balance there about how I can enable that engineer that, you know, has been doing that. Now they can write code much faster, they can oversee things, and that enterprise scaffolding in the middle allows us to actually create the right guardrails, create a safe system of work for them to go develop things in. And then it's also the feedback loop. So the other thing that we do with our ontology and our platforms is implicit and explicit feedback from users using it. So the OODA loop, that I create. And really that OODA loop allows our customers as they're doing workflows, they're giving feedback to agents. Now, can agents help them do more based on the feedback? So both explicitly saying, hey, that was wrong and this sucked, or I chose this option. Now, if you do that enough, agents can start to learn from that. So we actually store that in our ontology to allow it to scale. So it's really that human-centric process around AI. AI is not like— we shouldn't be thinking about AI for the sake of AI for AI. It's AI to enable humans to do more.

Yeah.

That's the frame.

OODA loop, observe, orient, decide, act. Yep. Right. I have a different question, but you can go.

If you were giving, if you had 30 minutes to give feedback to the AI labs, what are the kind of key areas? Let's say the frontier labs, right? Leading models. What are the kind of key areas that you would be focused on?

Yeah, I mean, I think when we think about the enterprise space, you know, we—

One, you're like, Don't compete with us.

No, I actually like— I think optionality is a good thing. Like, I am agnostic to where you store your data, where you store, what model you choose, what compute you use. So like, we can allow you to use any of that because the last thing that actually drives an outcome is replatforming, moving to another.

And that goes back to the on-prem culture, the secure cloud cultures, ITAR compliance. Like, this is in the DNA of the company.

And so how do we actually enable people where they are?

Yeah.

Instead of the focus on, oh, if you replatform everything to Palantir, everything will be great. And we're like, well, actually you've probably been replatforming for years. Can we enable what you have to go do these new things? So when we think about the model companies and it's how do we ensure that we can get the feedback loops around tool usage and—

Yeah, that's the kind of stuff I was wanting to get your point of view on is like, I'm sure you're getting into the nitty gritty with individual models, where they're spiky, where there's shortcomings, et cetera.

Yeah, so we actually just launched, I just put a YouTube video out last week on this new tool called Evolve. We talked about it in the halftime show where customers are using actually AI to help them understand which models. So maybe the meme around, hey, make it exist first and then make it good. Most of the time I see people building with agents, they're using the latest frontier model. Always. I just got it working.

Yeah.

And then all of a sudden the token maxing and everyone, everything else, and you're like, oh my gosh, I just blew through my whole budget. Yep. So we built a tool called Evolve that will actually go analyze the logs in production about how these models are operating, what people are doing with them, the architecture over it, and actually be able to swap out different models from different providers. Or hey, actually for most of this workflow, you can use this model that's older and actually without thinking and test time compute it, it's more deterministic.

Or even cached models.

Cash models. And then, or hey, if you actually just had this piece of data in the ontology, then you would eliminate all this and 50% of your cost. And so, you know, some of these customers— McCarthy talked about this at our halftime— you know, they were able to in 2 days eliminate 60% of their token cost by re-architecting, picking a different model and prompt tuning. So it's the combination of all those. The permutations get really hard, especially when it's in this probabilistic models. We have tools to do this in the deterministic world.

Prompt tuning. It's a— instead of don't make mistakes, it's okay to make some mistakes. Mistakes. If the mistake is going to cost just a little bit, I'm fine, because don't make any mistakes. That's going to cost me a fortune.

Well, there was some chatter yesterday around something a model was doing to be more efficient was talking in like this bad—

Oh, oh, caveman. Caveman prompting. Yeah, the caveman prompt method actually works.

How often are you working with a company that is having call it like a mini ChatGPT moment within their enterprise. And then they're just like, let's not tell anyone about this. Because I imagine like there's all these— there's clearly places where—

what does that mean? Their product is taking off like ChatGPT?

So they've found a way to apply AI in a way that is highly, highly effective and gives them an edge.

Oh, interesting. Uh, but yeah, like, like a theoretical, like, within like technology, how to transform.

Yeah. So So X people are very loud, right? Yeah. They're like, I just had this—

I'm using everything.

Yeah, I just had a product work for 30 hours on this thing. They'll talk about it. But if you're a Fortune 500 and you figure out how to do something, it's not like you want to like put your hand up and say like, guys, like I figured something out, right? Like secrets are valuable and these advancements and kind of breakthroughs are not going to be uniform.

The airline industry will never be the same. Your direct competitor copies you. Yeah.

And so, and so So part of why right now the meme is token maxing, and that's going to be an obvious area of debate. People are happy to go talk about it. CEOs might say, hey, let's stop doing this. But there has to be all these other pockets of interesting moments where we won't hear about them until they become standard operating procedure.

Or you see it in the— Earnings. Earnings and the economics piece, right? Yeah, yeah, yeah. So yes. Unfortunately, X is not the real world, you know, and there's a lot of grift and noise and, you know, podcasting, PMing and, you know, that kind of stuff that goes on. But I think in the real world, yes, there is the haves and have-nots. I mean, we were just talking about AIG, like when you can start to actually do the underwriting and, you know, have quotes back in hours or days instead of months on these highly complex enterprise, you know, kind of insurance agreements. If you don't have that, how are you ever going to compete? And so when we think about this of the N of 1, right, you know, those are the companies that we're going after and we see where there are those moments that are not public.

It's a competitive advantage. That's an interesting category because you can imagine AIG, you know, is, you know, working with a potential customer or renewing a policy and that customer is going and talking to all of AIG's competitors. And if AIG is able to turn around, you know, a quote or a policy in 24 hours and then it takes another player, you know, 2 weeks because it's, you know, complicated email and spreadsheets. So many, so many teams will just say like, hey, we, you know, you know, especially once you have 2 bids, you can basically say like, okay, there's that third, fourth, fifth, we'll kind of wait on those because we have, yeah. Uh, good option here.

Well, it builds trust. Yeah. The other piece here. So when you think, when you see people operating that with that level of efficiency, What else can you do? So I see this whether I'm doing SAP migrations, the least sexy thing you could talk about, but hey, if I can cut your SAP migration.

Let's give it up for the SAP migration.

Yeah, it's like the least exciting thing on paper, but actually if you're spending hundreds of millions, yeah, you guys get it, but hundreds of millions of dollars on a migration and we can cut it in half, that's a massive deal.

Back on the OODA loop, observe, orient, decide, act. On the observation side, what is the supply and demand imbalance for dashboards? Like, and what I mean by that is, is when you're working with a company, is there, is there more demand for dashboards? More people asking, hey, we need a dashboard for this, we need a dashboard for that, and you have to back people off and say, I don't know if the dashboard's right for this, like, you might just want to do an ad hoc analysis or actually go and see, uh, versus you're seeing so much opportunity that you're like, okay, we want to push dashboards out everywhere. Like, walk me through dashboarding right now because I've always been like, sort of like, oh, the too many dashboards. You build them and then no one looks at them.

Yeah, I want to kill all dashboards.

Okay, that's my perspective.

Dashboard, I mean, KPIs and dashboards should be a byproduct of operational applications where I'm making decisions. So we talk about the OODA loop. I have to actually act for things to hit the bottom line, be valuable.

In the actual application.

In the application. So as I need those things, and it's gonna inform a better decision, that's where I want those metrics. That should be a byproduct. If I go out with the goal of building a dashboard, it's gonna be the field of dreams again. No one shows up. And so yes, it should be, you're gonna have to build some of those things. The other side of this also is when you think about a data warehouse, like literally, I won't go too deep into this technical riff, but like, you know, Kimball and dimensional modeling was built in '96 for scaling databases and you're still modeling the same way in 2026. For your dashboard, your Tableau, whatever those things are. And like, that's not actually how the world works in rows and columns. You need complex things to model how the world really works. And that's what we think about the ontology, which means I can reuse it for an operational application, KPIs, agents, all in one single ontology, which makes it the compound effect where as I add things in, I'm now compounding with each individual decision I'm design— working with gets better and better and better for the next use cases I connect across my business.

Yeah. Is there an analogy there to just the deployment of AI tools currently? I'm just reflecting on the NoSQL boom and I don't know how strong this was. This is probably just like an online take, but this idea of like, why would you ever want a relational database? Why would you ever want a schema? Don't ever do a migration ever again. And the future looked like a win-win almost. Like I think Postgres installations probably grew and so did MongoDB and other non-relational databases. And people use Redis for things and they use all sorts of different tools. And we built and we stood on the shoulders of giants and we got more giants. And then, you know, that means full employment for you, obviously. But, but I'm wondering, like, as like, are you seeing glimmers of, of the AI tools eating into different pieces of the technical stacks, or is it all like yes and across the enterprises?

I think it's yes and in a couple of different things. There is when you think about the real world, it is not just rows and columns. You can't describe everything with measures and attributes. Yeah. And so it's actually multimodal. And so like we think about this in our ontology where you can have one semantic object that actually has a CAD file and an image, a CV model and tabular stuff in one semantic thing of a plant. Yeah. Which means I'm starting to talk in the language of my business. So being able to have the multimodal representation where in other places, oh, I have to have MongoDB and I have to have a SQL database here and I have to have an S3 bucket here. To put all of these different things to store them in ways. Well, we can do that all in the ontology, vectors, everything else. So that's really the goal around how do I model the real world, how it actually works, and make that transparent so you're not having to figure out which technology to put in a time series thing for sensors on an oil platform. Don't care, right?

And that's where we want to have the non-differentiated heavy lifting, like truly in the platform to remove the friction about getting stuff How common is it for a business with more than $100 million of revenue to have very little understanding of how their business actually works? Maybe they know the main thing, which is we make a product and try to sell it for more than it costs to deliver. But is some element of— how much can chaos and mystery be reduced effectively today? Because it feels like we're entering an era— like you go back, you know, 50 years and the level of like mystery in a large company would have been like is almost inconceivable today, right? Because you have different time zones, different offices, you know, no email, all that stuff. And now, like, mystery and chaos is probably reduced dramatically, but still there's companies that maybe before you start working with them, I'm curious what those look like.

Yeah, I mean, we work with a lot of different varieties of companies. I joke that a lot of times companies make money by accident. They don't actually know what their most profitable product is, and often they're trying to sell the thing that isn't actually the most profitable and actually not selling the thing that actually is profitable. And it comes back to how they've modeled the data to aggregate it up to KPIs and other metrics when you actually need to model at the finest grain how your business operates to get a true cost of goods sold, for example, or true cost to serve. Like, that's very complicated. It's very complex. So like, we really think about how do I embrace that complexity so that I can truly understand tactically at the edge, how do I do more of the things that are good and less of the bad? It's that simple. And those get peanut buttered across with KPIs and metrics. Metrics and people don't actually know how their businesses are up. And I can't tell you whether it's a $100 million company or a $50 billion company, how many times I see this, that they don't actually understand how they're making money at a fine grain.

Yeah.

Last question. Is there a world in the future where a company gets created, let's say on Stripe Atlas, and the first account they sign up for other than that is, let's say, a Palantir?

Ooh, yes, I would love that. And so we do have a Palantir for Builders program. We have small companies. There's people here that are built on top of it, two-person startups, you know, that are working in their attic in Canada. I mean, like, so it is literally any size company come work. There's a free dev tier. People can come build. You actually— there's actually a Shopify integration in Palantir. You can go hook up to your Shopify and pull in Palantir. Yeah, there are people doing this now. Are we always great at selling it or telling the story?

Sure.

No, but, but there are companies doing this. And I do think there's a day where it's going to be ubiquitous because I also think, you know, there's some, some guys here that have, you know, they— hey, my, my business is dying. I was, you know, I was down 10% negative margin on what I was selling. And through using Palantir, they watched our YouTube videos and they built it themselves and increased to a 9 or 10% positive margin in 3 months. That's great. And so like people can go do it. I think that's the great American story is like, how do we enable that? And I think I think we'll get there. Um, it might take a little time.

I love it. Well, thank you so much for taking this time.

Thank you.

Great to catch up.

Great to see you.

We'll talk soon. Uh, our next guest is joining in just 15 minutes. We're going to go back to the timeline. First, I'm going to tell you about Figma agents meet the canvas. Your AI agents can now create and modify Figma files with design system context.

It's so crazy how many companies Yeah, our— their whole strategy is like, we're gonna hire guys like Chad. Yeah. And they're gonna— they're gonna do stuff. Yeah, he is— he is—

he is the final boss of FTX. What— what— what drove the FTX meme? Was it— was it Palantir going public, or was it—

I think it was Palantir going parabolic.

Maybe. Maybe. Yeah, once—

yeah, because it just— it just—

there was like a chart, because before it was like, okay, yeah, successful company, but like no one really knows where the valuation's going. Now it's like My uncle just told me that he made a bunch of money, and so I gotta pay attention to this.

Yeah, well, that, but also they had been banging the FTE drum and getting the consulting activity.

Yeah, but people had earplugs in to the banging of the drum, and the earplugs came out.

Yeah, but when they were a $10 to $20 billion company, a lot of people could still convince themselves they were right. It's just a consulting business.

Yeah, yeah, yeah, exactly.

But now it's harder, harder to ignore. We covered this very briefly, but we are very excited for Joe Rogan to be hosting his, you know, it's rumored this is a rumored leak. It is not confirmed by any means yet, but, but I like the sound of it.

It would be. It's a very different direction.

This was a good post. I want to bring it up. Buko Capital says it's really incredible the absolute AI garbage in all caps that people are comfortable sending to their coworkers and bosses. There's a good chance productivity will actually decrease as adoption increases because everyone is busy wading through AI slop. I don't think— I don't think it'll actually— I don't think it'll actually get there. But I have had— I have had moments over the last month where somebody has sent me, you know, a deck for their company or materials, and I can tell that that, uh, 90% of the work that went into it was on prompting. Yeah. And, uh, I have a very, like, visceral reaction toward it, especially for, like, early-stage companies where, uh, ideas and the way in which you go about doing things matter so much that, uh, it's almost like, you know, painting this initial vision and things like your, your go-to-market product differentiation, why you'll actually win. Like, use AI to make your team slide. That's great, right? Just taking like a set of facts and making it look good, right? You're giving somebody a bio, something like that. But I just remember I got this deck, I was clicking through it, and I very respectfully said, like, go and like, do this yourself. Yeah, because just because you've made something that, that looks like a deck. Yeah. But you didn't do the sort of like fundamental work to actually present this in a way. If you looked at each slide individually.

Yeah. Your eyes kind of glaze over.

Yeah.

And you just sort of like lose focus. Yeah.

Yeah. It's like it would have been more— it would have been more compelling to actually just have a bulleted list of like problems I mean, a lot of times you can just—

obviously, send me the prompt because I can instantiate it in my head. I can imagine the rest of the paragraphs. I have the context window preloaded. Yeah, uh, for myself. Yeah, uh, we should talk about the new Audi, the NuVolare. Is this real? Motor1, this seems real.

It's real.

Uh, it's a big deal. It's the— it's the brand's first supercar since the R8. Twin-turbocharged 4-liter V8 hybrid. 217 mph top speed. That is 10% faster than a Cayenne Turbo GT. What is the Cayenne Turbo GT market doing right now? Is it tanking? Depreciation must be just through the roof on this news because you have a car that's 10% faster. And so why— everyone was going to be rotating out.

I mean, I, I think, I think the new Velar—

it's a really cool design. It's a really cool design.

Feels like somewhat Cybertruck, cyberpunk-y, futuristic.

I don't know. It just checks the box for like the next supercar for me. And in a way that the—

Ben says it can't touch the R8.

Oh, can't touch the R8. Okay. Okay. Well, it goes 0 to 60 in 2.6 seconds. Well, almost 1,000 horsepower. Let me tell you about Cisco. Critical infrastructure for the AI era unlocks seamless real-time experiences and new value with Cisco. And our next guest, Sam Berry, is here from the USDA. Welcome to the show. What's going on? How are you doing? Very good. Good to meet you.

Great to meet you.

Thank you so much for coming on down. Let's throw this on and just like that. Cool. On the left side.

So good.

Introduce yourself a little bit. Tell us about yourself.

All right. Yeah, my name is Sam Berry. I am proud to be working at the USDA.

What do you do there right now?

I'm the chief— Nominative determinism. Do you know about nominative determinism?

No.

It's the idea that a person's name could possibly influence or, or, uh, the, the— but, but berry and working at the Department of Agriculture is like pretty perfect.

Yeah, no, it's incredible. Actually, my, uh, the berries came over here from France in, um, like 1640. Whoa. So we've been here for a long time.

That's crazy.

And, uh, it was all farmers.

Yeah, there you go.

Yeah, yeah, it was like all farmers until my grandpa. Okay. Then he became a, uh, materials engineer actually and worked on jet engines.

Okay.

And so then his sons became engineers. My dad became an engineer, and I was an engineer. So we're kind of trying to bring the two together.

Now you're going back to the USDA. Yeah. What is the shape of the USDA? Like, what is the shape of the organization headquarters? Do you go in the office? Is this, you know, US? You think just America, international footprint? Like, do you travel for work? What's it like working there?

Well, actually, it'd be kind of interesting to ask you what you think. Like, what are the things that you think USDA does?

They grade the milk in the States.

Yeah.

Okay. That's what I think about it. So I imagine that at some point farmers send the cows to you and you kind of inspect them and say, this is a good cow. Is that what happens?

I don't know. There's inspectors. There's a whole area that does that.

I imagine there's a series of certifications, but what else is happening?

So all kinds of stuff. Do you know that food stamps?

Yeah.

SNAP is inside of USDA.

Oh, I didn't know that.

I didn't know that either. I figured it was like HHS or something. But yeah, it's in USDA. Yeah. So that's $100 billion a year. It's kind of a big deal. Yeah. So we do— we have SNAP that's in the Food Nutrition Service. Yeah. Forest Service is inside of USDA. Okay. Like crazy.

Yeah.

And then FPAC is like what you would really think that USDA— it's like the farmer-facing, like, okay, where farm programs are, where they do acreage reporting, like the stuff I talked about today. Got it. Then there's Rural Development. RD, which is like loans, like a bank basically. Do loans for all kinds of things.

Okay.

Actually, in some of the reviews I came in on Doge and there's like beachfront hotels that are being funded out of RD. So there's like a lot of things that need to be cleaned up.

Okay.

Yeah.

And then there's like food inspection service and then there's actually a huge scientific arm.

Yeah.

That's inside of—

that makes sense. Testing things and yeah, like labs advancing different pesticides.

So things that I mean, I actually become very passionate about it because I certainly didn't have an appreciation for— I thought the same thing. It's like grading meat.

Yeah. Milk, you know. Yeah.

But like, we are so uniquely positioned as a country because of the fact that we can feed ourselves. Yeah. And like, that is not the case for a lot of, a lot of countries.

Yeah. Isn't America basically a net exporter of food too? You hear about this in the China debate all the time. Will they buy XYZ products from us as retaliation. And yeah, you just don't think about it.

But yeah, so like China can like minimally feed itself, like bare minimum, like keep itself alive. Yeah. But you know, they're getting like, like we just did a big deal with them to move a bunch of beef over there.

Yeah.

And kind of got some negative press on that. So it's important to know it's— I forget exactly what it's called, but it's like the parts of the cow that we don't eat here. So it's a little misleading to say like the amount that we're sending over there.

Also, all these trade deals are like very complex. And there's like 6 different moving parts. We get batteries or they get the chips and like, yeah, these are always like, you know, 7-part negotiations. It's hard to look at anyone in isolation.

But I mean, I think it's a little surprising that agri— like food is actually part of that. I mean, in America, in warfare, like agriculture, the food supply is usually hit before anything like kinetic even happens, you know, and then before even the world knows that it's warfare. Oh, you know, okay. Because you can do that and you can do things to impact a nation's food supply in the future. And so agriculture is a really big deal. Sure. Really important. So all this to tie back to, I wanted to talk about the labs because this is a whole area inside of USDA, but we do all of these things like invest in figuring out— so personally, I try and avoid GMOs and we eat, we drink raw milk and we get our meat from a local farm. But GMOs are actually really important because if we were hit with some kind of adverse event or something, we needed to create corn that could survive a drought better. Like, we have the science and the research to be able to do that. Got it. And it's a huge edge that we have, like, geopolitically.

Interesting.

Yeah, yeah. Talk about, uh, over the years I've read so many stories of, you know, this, uh, this insect has been detected in, you know, some region of the U.S. and there's speculation on is it, um, kind of foreign interference, things like that. Is that in USDA domain, is trying to help monitor and track and make sure that— Pests. Yeah, pests. Pests are obviously naturally occurring, right? They flourish for their own reasons, or there can be some sort of malicious intent as well. Is that in your guys'—

Yeah, because they're not necessarily naturally occurring, right? And so one that we have going on right now, and I'm not saying this one's not naturally occurring, but the New World screwworm. Yeah, that's coming up through Mexico. So our secretary, which, by the way, I couldn't say enough good things about Secretary Rollins. I mean, she's incredible. Just an actual, like, genuine good— like, it's unbelievable what she's able to accomplish. But New World screwworm is something that's falling in USDA's, you know, responsibilities. And this is like a parasite, basically, that's coming up through Mexico. And it's like a flesh-eating parasite. So it's like really hardcore. Yeah. So we're developing a lab.

You have flesh?

No. All sorts? No. But, you know, I don't think you want to be around it. But no, it's for like cattle mostly is what it impacts. And so we're developing a lab and like sterilizing flies, which again, like personally, I don't really like any of this stuff, but it's better to be doing this and be able to protect our nation than like if we let this just come and flourish in our country, I mean, it'd be very detrimental.

Yeah. So I'll have to go back and if it's a necessary technique that needs to be harnessed, it needs to be harnessed securely and it needs to be harnessed with the right teams in place to make sure that whatever's rolled out is rolled out effectively and safely.

Yeah. I think it's just so important. There's so much farther we can go with technology, but we have so much right now and so many people are just blackpilled.

Right?

And I think it's important. I think you should be like blackpilled on certain things, but you should probably take a lot of pills. Like, you should be redpilled and blackpilled and whitepilled at the same time. Okay. Uh, because good take. Like, we have a long way to go, and when we're just like sitting feeling sorry for ourselves, like, it's not a good position to be in. Like, this is the most incredible country on Earth, and other countries are advancing though. You know, our edge is like— our edge doesn't come for free.

No, we got to work at it.

We got to keep pushing at these things. But when we do this, like, when there's a parasite that's coming into our country and we're able to just use biology to combat it. That's incredible that our country can do that.

Talk about these more SMB-scale farmers and their approach to technology. I think a lot of people would be surprised at how much these individuals, at least from what I've experienced, are happy to lean into technology. I met a group in Texas that had, developed— this was years ago, so pre-AI boom— developed their own SaaS product to help manage their operations, like a tool that they had built by discovering problems that they had on their property. And I just thought that was really fascinating and cool at the time because I think Silicon Valley would have maybe some expectation that there might be an aversion to that until you get into the more like enterprise-grade scale.

Yeah, I think it's a really important topic because you're essentially talking about like democratizing access to technology, right?

Yeah.

And certainly with like AI becoming so much more widely available, that was a big step forward. But I mean, and this is a big point that's being hit on at this conference and what Palantir is really focusing on is those LLMs become useless if they're not, if you're not deploying them in the right way with the right like data boundaries, right? So I think that's something that we're seeing even in our universities. We do a lot of university research and all the kids or whatever, the university students, they're wanting to do experiments with LLMs and do meat grading, better meat grading, because that's something that can happen at the farms. And if you can make that automated, then our ability to produce beef is greatly impacted. But there's a major issue in secession planning. Right now for farms, right? Like, this is a big thing that's happening. Like, the farmer generation is getting very old and kids don't want to go and run the farm.

A lot of them went to big cities and jobs and white-collar work and stuff.

So, you know, this is a big thing that is H-2A. Yeah. You know, these H-2A visas where like a lot of the farmers are actually still saying like, we need the help from, you know, we need immigrants to come and help us. And, you know, the best way that we can solve that is through automation. So I think that That's something I would love to see USDA do more of, or, you know, uh, it's something that needs to be answered. I don't have an answer for you right now. Yeah, but in order for us to continue to, you know, remain self-sufficient and providing food—

whenever you have a dwindling workforce, increasing the leverage and productivity of the existing workforce allows you to maintain overall aggregate productivity. This is general technological leverage, so it makes a ton of sense.

Do you know anybody that's becoming a farmer?

Uh, well, we know some folks. We've had a number of, uh, entrepreneurs on the show who are Getting into ag tech. Yeah. Building, uh, we've had the founder of the LaserWeeder that, uh, uses, uh, a lot of people don't like pesticides, but they don't mind if, uh, if a pest is zapped with a laser because that's just heat that's being transferred to the particular plant right there. And the tomato plant continues flourishing. So he uses just cameras and lasers. Very cool sort of modern solution to something that people have had a lot of fear around, around different pesticides.

Yeah. We had a fruit, fruit picking robotics company. Yeah.

Orchard as well. But mostly from tech side, usually with some family lineage, sort of returning to the roots or tapping into their networks to go back. But I mean, truthfully, I don't know that many people that I grew up with. I mean, I grew up in LA, so not much farming activity. I knew one family that had an avocado farm.

I mean, it would be super based to be a large-scale farmer. More people should do it.

Maybe you could be the Alex Hormozy of farming.

Yeah.

No, for real.

I mean, you can. So USDA, one of the great things that USDA does is you can get financial assistance, like you get big time, like big time loans. Yeah. You have to go through the process. And they were actually doing a loan modernization effort right now, trying to make that better. But like USDA will fund it for you. You got to pay it back, but you can like get the interest rate very, very low rate subsidized.

Yeah.

Yeah. I mean, one of our administrators at USDA, he like pulled up his phone one day and he's like, look, it's a planting day for me. And it was this John Deere app.

Wow.

It's like the most advanced deal. Like he had all these tractors going and there's still people sitting in the tractors, but it's to the point where it basically could be fully automated. So I mean, you can get yourself a couple thousand acres and just start growing corn or wheat or cotton, like cotton, and then whatever.

Talk about data collection. I feel like data is the lifeblood of any decision-making, any OODA loop, anything related to Palantir, USDA. And I'm wondering about, like you mentioned that screwworm, you got to track that thing. It shows up on some cattle rancher's farm and they're detecting it or they're seeing symptoms. Maybe they know roughly what percentage of the herd is affected, but how do they actually get that information to you? Are they going to usda.gov/reportincident, or are you pulling things from their filings? Like, how, how do you want that to evolve? I imagine that, that with more AI and technology, we're— it's only as good as the data that we can actually put into the system. So just broadly, data collection, where is that going? These days?

Well, if you don't mind, instead of screwing, I'd like to focus on SNAP.

That's great.

Yeah. So SNAP is funded by the federal government. Yeah, but it's administered by the states. Okay. So when it comes to— so something that we're doing right now, and it was one of the first things that our secretary did, like on our first day, was she did a data call to all the states that, you know, we want all of your SNAP data to understand how— because it's our responsibility as the funder of this program to understand the integrity, like to verify the integrity of the program.

Yeah.

So we put a request out there, but it has to come from every single state.

Yeah.

And a lot of the state programs, they're not technical or they've got contractors that, you know, it's just a difficult thing to get us the data. But then there's also a bunch of states that are just not complying, you know, for whatever reason, which it shouldn't be a problem. I don't understand what the problem is, but the importance of— so that program, that's $100 billion taxpayer dollars a year. Like, that's pretty substantial. That's an area where we really want to have all angles of the data available so that we can deploy AI and become really smart in detecting fraud. We want to get it to the point where if somebody is committing SNAP fraud, we should be able to— it's like your card, right?

If you—

if somebody stole your card and did a transaction that wasn't recognized, like your card shut off.

Yep.

Right. Yep. So we want to get to the point where we're very intelligent and we're confident enough in the system that we can do that. When there's fraud detected, it's off immediately because it's an important program. You know, we want to be able to support people that can't support themselves. But it's, it's not arguable that there's a massive amount of fraud in there. I mean, even the, the organization itself does like an audit every year and they're at like, there's 12% improper payments. Improper payments is kind of a bad word.

So $12 billion a year.

Yeah, right. And that's just like kind of based on—

that's money that, that's, that could actually be going towards the intent of the program. Yeah. Which is to provide food to people that otherwise wouldn't be able to get it.

And there's other, you know, you could like grok how SNAP has been used to fund like international crime organizations and like terrorist groups and everything. So it's, it's being exploited at a huge level. And I mean, it's something that our secretary has prioritized, but that's probably our biggest FPAC. What I talked about today is like our most complex system of data. But the SNAP challenge is like the biggest or like the SNAP environment is probably the biggest challenge on the data front.

What's next for you? Are you making a career out of this or are you going to go be a farmer?

Hopefully both.

Okay.

Yeah. Yeah. I mean, yeah, I've got some farmland.

Yeah, you do.

Trying to convert it. It's like woods right now, but—

What state? Where is that?

In Virginia. So actually, when I lived in Michigan, we had like a little bit of a farm. We had some goats and sheep and and a bunch of chickens and ducks. You don't ever want to get— you don't want to get ducks. You don't want to get goats. Ducks are like really savage, actually.

Yeah.

Like a chicken sleeps, you know? So like it's got a normal cycle. Like at nighttime it goes into the coop and it sleeps.

Ducks don't sleep.

No, ducks do not sleep.

They like— in our house was kind of this like really unique house. So the windows were like on the ground and the ducks would come and just stare at us in the window. Oh, they're savage. They just like, they sleep for like 10 minutes at a time. So they'll just like waddle around and then sleep for 10 minutes. You have to have the right balance of female and male ducks. Okay.

Otherwise it's like, that's really ugly. Yeah. Chickens are a lot. I grew up with chickens and most of the time they're cool. My dad would build these sort of like complex contraptions to automate the opening and closure.

Oh, interesting.

So he would use like irrigation to, on a timer to fill a bucket, which would lift their, would lift it up.

Yeah.

Yeah. Interesting. But then I still, core memories as a kid was waking up, my dad would yell like, "There's a fox in the coop!" And then we'd be running out.

Really? You'd have to run?

It would be like game on. Yeah, yeah. Or you'd get skunks in there. Yeah, or coyotes maybe. Yeah, everybody would get up and try to go deal with it.

That's got to be satisfying. That's way more satisfying than some software bug.

There's so much fear and doom and and blackpilling around data centers. I wanted to hear from you how you're— I imagine your role is to be an advocate for farmers as well on water supplies, things like that. California went through probably many, many really rough years from a water supply and a water scarcity standpoint. Thankfully, had a lot of rains over the last few years, but how are you working with farmers, or what is the situation around the tension between a lot of farmland could also be great land for data centers, right? And there's been some pretty high-profile stories where farmers either sold their land, but from your side, you're trying to make sure that we have, can produce an abundance of food from a from a national security standpoint. So how are you guys thinking about that balance?

Yeah, I mean, I think the best solution is putting the data centers in space, you know, like, which is totally led by Elon, and people are jumping on that train. But it's going to be a couple of years, it sounds like, before they're to that point. We're actually— USDA is pursuing a partnership with SpaceX, and that part isn't ready yet. We don't really have a need for that. But it's— there's a partnership on the technical side, but there's also just on the, like, conceptual side of the fact that, like, we're aligned because we do care about conservation. You know, there was a lot of green stuff that was like, you know, it's not stuff that we care about, but we do care about conserving our land and putting data centers in space just makes a ton of sense. But that being a couple of years out. So for today, you know, I'm actually pretty passionate about this because in my hometown of Saline, Michigan, it's like small town, mostly farmland. They're putting a data center in there and it's like, you know, 30 miles from Detroit and Flint and like all these very industrialized areas. And so it's very confusing to me why we wouldn't be putting these data centers in there, like struggling areas. Detroit's doing all right, but like Flint struggling big time. Like, why not put a data center there where there's already the infrastructure? There's like, it's already developed land, but instead it's like taking these small townships and plopping them in the middle and the people don't really like it. Now the boards seem to like it for some reason, the councils. So I don't know what's up with that, but it doesn't align with what creates—

it creates creates a massive amount of tax revenue that can be used to fund a bunch of other programs.

Yeah, but it's got to actually flow back to the people who are in the town. And I think that there's like a disconnect there sometimes.

I actually— this is kind of outside, but something that I do think is, uh, probably going to happen is, you know, there was this big shift to go to the cloud, right? It's like everybody kind of had their own servers, you know, it's on-prem, and now we're in the cloud. And it's like, really, you just took you like moved it across the street, right? And now that people are becoming more aware of like what that means and when it's like, oh, my data's in AWS or, you know, it's like, and maybe this is a global company and how much can I really trust this company that there's going to be a shift back to caring, actually, actually caring about where your data is living. So I think a good business opportunity would be— I think there's a world where there's a culture that comes up around data centers because like me personally, like I want to build like my house is like a, like I'll have a kill switch for my Wi-Fi and then like we've got the data in the basement and got your raw milk supply. Got raw milk. No, like we're ready to go. I mean, I was ready to go off the grid before I came and joined the government. This is a much better option. But still, like, I care about my data. I don't really want to use YouTube Music anymore for my music because now my recommendations are getting worse and you're like very beholden to that. It's like I could very easily just have the music, buy my music. And write a simple program to like make my recommendations. And it would be way better because there are certain artists that are not getting recommended because they're not, you know, prioritized behind the scenes or something. So, but not everybody's going to want to manage their own servers, right?

Jetson just announced a data center that bolts onto the side of your house.

Oh, that's sweet.

And I mean, and there's more stuff that's coming that way. I mean, people are doing the Mac Minis, can't really do the Frontier AI on the Mac Mini just now, but in a few years, you know, the, DGX desktops, like it's all coming and I think it will be more of an option.

So just to like kind of wrap this up. So there's this or this like topic, the one of the things that USDA does is we pay 600,000 federal employees. So like we pay Secret Service, we pay DHS, we pay— it's like, it's like a thing inside of USDA. Interesting. And so the payroll system that does that is a mainframe. And people literally explained it to me like this thing has a personality, like you have to like You can't touch it the wrong way. You have to— it has to have like the right environment to work. And it like all these things. I mean, like a dozen people came to me, told me all these things. So then I went and visited it. I was like really excited to, you know, encounter this being mainframe. And it's like a 5-year-old brand new like IBM server. You know, it's just like it's not— there's no tapes. There's not like a team of people. You're underwhelmed. Yeah, it's like, you know, it's like this big.

Okay.

But I was expecting like a— yeah, like a small micro data center or something. Exactly. Yeah.

So totally modern. And it like, I like formed this connection with it and I was like, we have had so many conversations about you. And I just thought that like, this is potentially a future where it's like a data center, a coffee shop, you know, like people might want their data to be hosted in a place that's like aligned with their views.

Sure, sure.

Yeah.

You know, because it's like, I can trust, like, I don't want this in my house, but I can like trust this like cool company, local company that my data lives there because I don't need it distributed across the globe. It's like, yeah, I'm here.

No, that makes sense.

That's interesting. Country intelligence.

Yeah, yeah, we talked about this. This is the future. I love it. Uh, anyways, so much for coming on the show.

Great to meet you. Thanks for having us. Thanks for doing this work.

Have a great rest of your day.

Thank you.

We'll talk to you soon.

We will wrap up the show. Yeah, thank you for tuning in with us today, folks. We will be back on Monday. Yes, and we look forward to it.

Some business to do tomorrow, but See you Monday. Leave us 5 stars on Apple Podcasts and Spotify. Sign up for the newsletter, tbpn.com, and have a wonderful weekend.

We'll see you.

Love you.

Goodbye.

Goodbye.